Hypersensitivity Reactions to Radiocontrast Media: The Role of Complement Activation Janos Szebeni, MD, PhD

Address Department of Membrane Biochemistry, Walter Reed Army Institute of Research, 501 Robert Grant Avenue, Silver Spring, MD 20910, USA. E-mail: [email protected] Current Allergy and Asthma Reports 2004, 4:25–30 Current Science Inc. ISSN 1529-7322 Copyright © 2004 by Current Science Inc.

Although intravenous use of radiocontrast media (RCM) for a variety of radiographic procedures is generally safe, clinically significant acute hypersensitivity reactions still occur in a significant percentage of patients. The mechanism of these anaphylactoid, or “pseudoallergic,” reactions is complex, involving complement activation, direct degranulation of mast cells and basophils, and modulation of enzymes and proteolytic cascades in plasma. In this review, basic information on different RCMs and their reactogenicity is summarized and updated, and the prevalence, pathomechanism, prediction, prevention, treatment, and economic impact of hypersensitivity reactions are discussed. Particular attention is paid to the in vitro and in vivo evidence supporting complement activation as an underlying cause of RCM reactions.

Introduction Radiographic procedures using water-soluble radiocontrast media (RCM) represent routine medical practice today in all developed countries. In the United States, for example, more than 10 million tests are performed yearly with various RCM [1•,2••]. Nevertheless, intravenous (IV) use of RCM still carries a significant risk for hypersensitivity reactions (HSRs) manifested in a broad range of cutaneous, gastrointestinal, cardiopulmonary, and other symptoms (Table 1). These reactions, also referred to as “RCM reactions,” are usually not IgE-mediated and, therefore, belong to the category of “pseudoallergic” or anaphylactoid reactions. Nevertheless, true IgE-mediated acute reactions and nonacute, delayed-type HSRs to RCM have also been reported [3,4]. The symptoms listed in Table 1 are similar with all RCM and essentially correspond to the acute allergic or pseudoallergic reactions caused by numerous drugs, including antimicrobial and antineoplastic agents, non-

steroidal anti-inflammatory drugs (NSAIDs), certain analgesics, liposomal drugs, and water-insoluble drugs that are dissolved with the help of Complement (C)-activating emulgents (eg, Cremophor EL) [5,6••]. However, some symptoms are more frequent with RCM compared with other reactogenic drugs, and there are also differences among different opaque agents in the spectrum of the hypersensitivity symptoms they cause. Radiocontrast media can be classified according to their iodine content, osmolarity (hyper-, low-, and iso-osmolar), level of ionization (ionic and nonionic), and level of polymerization (monomeric and dimeric). Earlier RCM preparations were mono- or diiodinated compounds (eg, pyridines); however, the structure of current RCM is based on fully substituted benzoic acid with three iodines at positions 2, 4, and 6 on the benzene ring. The ionic monomers are sodium or meglumine salts of the anionic triiodinated benzene ring; these are the high osmolarity RCMs (>1400 mOsm/kg). Their dimerized derivatives have lower osmolarity (eg, 600 mOsm/kg), and further derivatization with hydroxyl groups or other hydrophilic conjugates results in nonionic, lower-osmolarity (500–700 mOsm/kg) or isoosmotic RCM [2••]. Hypersensitivity reactions have been a concern since the first organic, iodinated compound was used for IV pyelography in 1928 by Swick, an American urologist [7••]. Subsequent development of new-generation RCM, characterized by a transition from hyperosmolar to iso-osmolar, ionic to nonionic, and monomeric to dimeric preparations, was driven by the need to reduce the reactogenicity while maintaining or improving the opacity of products. Therefore, the initial hyperosmolar, ionic, monomeric class of RCM (eg, diatrizoate, iothalamate) was outdated in the early 1980s by lower-osmolar, ionic, dimeric RCM (eg, ioxaglate). Progress then continued with the introduction of nonionic, monomeric, lower-osmolar RCM (iopamidol, iohexol, ioversol, iopromide, ioxilan), and finally, in 1996, the US Food and Drug Administration (FDA) approved iodixanol, the nonionic, dimeric, iso-osmolar RCM [2••]. Table 2 lists the bestknown RCM, with specification of their above-discussed physicochemical properties. It should be noted that in addition to categorization according to molecular and physico-

26

Anaphylaxis and Drug Allergy

Table 1. Symptoms of RCM reactions Category

Symptom

Common allergic

Angioedema, chill, choking, conjunctivitis, coughing, diaphoresis, edema, fever, headache, nausea, rhinitis, sneezing, wheezing Erythema, urticaria, flushing, rash, dermatitis, skin eruptions, pruritus Emesis Hypertension, hypotension, cardiac arrhythmias, cardiac arrest Tachypnea, dyspnea, bronchospasm, asthma attack, cyanosis, pulmonary infiltration Chest, low back and lumbar pain, confusion, sensation of warmth,* tingling, panic of imminent death Generalized maculopapular exanthema

Cutaneous Gastrointestinal Cardiovascular and hemodynamic Respiratory Vegetative/neural Delayed (days)

*Particularly characteristic of RCM reactions. RCM—radiocontrast media.

Table 2. Physicochemical properties of RCM that influence hypersensitivity reactions Generic name

Osmolarity

Ionization

Polymerization

Diatrizoate Iodixanol Iohexol Iopamidol Iopromide Iothalamate Ioversol Ioxaglate Ioxilan

HyperIsoLowLowLowHyperLowLowLow-

Ionic Nonionic Nonionic Nonionic Nonionic Ionic Nonionic Ionic Nonionic

Monomeric Dimeric Monomeric Monomeric Monomeric Monomeric Monomeric Dimeric Monomeric

RCM—radiocontrast media. Adapted from Hong et al. [2••].

chemical properties, earlier classifications also considered the severity and clinical manifestations of HSRs (nonreactogenic, mildly or severely reactogenic RCM; anaphylactoid, vasomotor RCM); the diagnostic procedure for which they were used (eg, myelography, angiography, arteriography, venography, urography, arthrography, CT); and the pathogenesis of HSRs (C activation, IgE-binding, direct degranulation of mast cells and basophils) [8–10]. The goal of this review is to briefly summarize and update basic information on different RCM and their reactogenicity, discussing the pathomechanism, prediction, prevention, treatment, and economic impact of HSRs. C activation is one of the most thoroughly studied causes of RCM reactions. In this review, particular attention is devoted to the in vitro and clinical studies attesting to this mechanism.

Prevalence and Critical Factors Influencing Radiocontrast Media Reactions In addition to the physicochemical properties of RCM, the main factors influencing the rate and extent of RCM reactions include the amount of agent infused; route (site) and speed of administration; premedication status of patients; and criteria for positive reaction. With all these variables, it is perhaps not surprising that the reported incidence rate of different types of RCM reactions span a range as broad as

0.0004% to 60%. According to comprehensive estimates applied for all kinds of symptoms with all procedures and all types of RCM, the overall incidence rate of RCM reactions is 2.1% to 12.7% [2••]. The frequency of severe anaphylactoid reactions is estimated to be up to 1% to 2% [2••], whereas that of nonsevere, cutaneous, vasomotor, pulmonary, cardiovascular, or gastrointestinal symptoms is in the 5% to 8% range [1•]. However, to illustrate the critical dependence of these statistics on details, life-threatening RCM reactions during coronary angiography occur only in 0.0004% to 0.002% of patients [11]. As for the relationship between RCM properties and HSRs, it has been proposed in several studies that nonionic, low-osmolarity RCM (LO-RCM) are safer than ionic, highosmolarity agents (HO-RCM) [2••,11–14]. Barrett et al. [13], for example, enrolled 1856 patients to compare the reactogenicity of LO- versus HO-RCM used in cardiac angiography. The overall incidence of HSRs requiring treatment was 9% versus 29% in these groups, respectively, and hemodynamic deterioration and severe or prolonged reactions also occurred more frequently in the HO group (2.9 % vs 0.8%) [13].

Pathomechanism The pathogenesis of RCM-related immediate HSRs cannot be explained by a unique mechanism. As illustrated by the

Hypersensitivity Reactions to Radiocontrast Media • Szebeni

27

and basophils for immediate release of preformed allergic mediators, such as PAF, histamine, and tryptase, but not those that are de novo synthesized in response to activation of these cells via IgE or C5a receptors (LTB1, LTC4, LTD4, LTE4, TXA2, and PGD2). The release reaction was shown to be Ca++dependent and similar to the effects of Ca++ ionophores; however, RCM (diatrizoate) had a more complex effect on cell function as it rendered basophils refractory to subsequent release reactions [17]. Another important detail concerning the secretory response of mast cells and basophils is that these cells display differential response to different RCM or other stimuli. For example, mast cells from the skin might not respond to certain RCM, whereas pulmonary and cardiac mast cells are triggered for strong release reaction [18]. Likewise, mannitol, an osmotic stimulus, might induce the release of histamine from human basophils, but to a lesser extent from mast cells [18].

Complement Activation As an Underlying Cause of Radiocontrast Media Reactions

Figure 1. Mechanisms of RCM reactions. This schematic illustrates the various triggering mechanisms and secretory products of mast cells and basophils during RCM reactions. C—complement; C1INH—C1-esterase inhibitor; LT—leukotriene; PAF—platelet-aggregating factor; PG— prostaglandin; RCM—radiocontrast media; TXA—thromboxane.

scheme in Figure 1, major triggering factors implicated are 1) activation of the C system with consequent activation of mast cells and basophils via C5a and C3a receptors; 2) activation of mast cells and basophils via IgE receptors; 3) direct secretory effects on mast cells and basophils due to local changes in osmolarity and ion (mainly sodium and calcium) concentrations; and 4) activation of plasma proteolytic systems, in particular the coagulation and kininkallikrein systems, entailing additional C activation by various crossover amplifications and depletion of C1-esterase inhibitor (C1INH) [6••,15]. As for the effector arm, the best known secondary mediators of anaphylaxis are histamine, tryptase, platelet-activing factor (PAF), leukotriene (LT)B2, LTB4 , LTC4, LTD 4, LTE4 , thromboxane (TX)A2, prostaglandin (PG)D 2 and TXD4 [6••,8–10,12] (Fig.1). The pathophysiologic effects of the best-known mediator, histamine, are mediated both by H1 and H2 receptors. Specifically, H1 receptors contribute to the cardiovascular and cutaneous symptoms via inducing vasoconstriction and vascular leakage, whereas H2 receptors cause vasodilation, increased heart rate and pulse pressure, and basophil degranulation via rising cellular cyclic AMP levels [16]. Direct secretory effects on mast cells and basophils involves non–receptor-mediated stimulation of mast cells

Activation of the C system as the main underlying cause of RCM reactions has been studied since the 1970s via numerous in vitro, in vivo, and clinical studies. Nevertheless, the picture that has emerged is far from consistent, and the question of whether C activation is the major cause of HSRs, a contributing factor, or only an epiphenomenon, is still open. In reviewing the in vitro data, it is clear that RCMs have several different effects, both within the C cascade and in its regulatory system, and that charge, viscosity, iodine number, hydrophilicity, and osmotic pressure are all critical variables in these interactions [10,19]. C activation by RCM was demonstrated to proceed via both the classic and the alternative pathways [10,19], as well as via unusual mechanisms, such as: 1) nonspecific, nonsequential cleavage of C proteins [20]; 2) suppression of natural inhibitors of C, such as factor H and I [10]; and 3) direct action on the thioester bonds of C4 and C3. However, another activation mechanism implicated is an electrolyte imbalance in serum [21]. Some of the in vitro studies contradict theories indicating C activation as the underlying cause of HSRs, or, at least, the authors interpret their data as contradicting the C concept. Vik et al. [21], for example, found no increase in serum SC5b-9 levels following the addition of iodixanol to serum, leading to the conclusion that it was unlikely that C activation was responsible for iodixanol-induced anaphylactoid reactions in humans. Lieberman [10] suggested that depletion of C proteins in human serum was not due to true C activation but rather to nonspecific binding of C proteins to RCM, without anaphylatoxin liberation. At the extreme, Mikkonen et al. [22] suggested inhibition of C activation by RCM on the basis that iohexol, ioxaglate, iodixanol, and meglumine amidotriz solutions effectively blocked insulin-induced generation of C3a-desArg in normal human serum. The authors claimed that these mol-

28

Anaphylaxis and Drug Allergy

ecules, particularly the ionic ones, inhibited the binding of factor B to surface-associated C3b, interfering with alternative-pathway amplification. Considering these C activation–opposing data and the technical intricacies and possible conceptual pitfalls in interpreting C assays [23], many data in this field require a close scrutiny with regard to conclusiveness and biologic relevance. Several animal studies attested to a causal role of C activation in RCM reactions. Lasser et al. [24] reported severe “idiosyncratic” responses in dogs to the injection of sodium iothalamate, manifested in vomiting, hypotension, and hyperreflexia. The authors found significant depletion of C during the symptoms, suggesting that C activation was causally involved in the reaction. In further dog studies by Lang et al. [25], serial daily injections of RCM (metrizamide, iothalamate, diatrizoate, acetrizoate, iodipamide, and iopanoate) caused substantial declines of serum C over several days. Napolovlu et al. [19] proved that various RCM in the 0.5- to 2.0-g iodine/kg range activated the C system in rats via the alternative pathway, with efficacy in the following order: triombrast > hexabrics > ultravist ≥ melitrast = omnipac. Human data on C activation during RCM reactions include case reports—for example—on a severe anaphylactoid reaction to a pyelographic RCM, manifested in a precipitous fall in plasma hemolytic C, C3, C4, and C1INH levels, with a rise of C3 conversion products [26]. This patient also developed consumption coagulopathy [26]. Vandenplas et al. [27] reported the case of a 29-year-old woman who developed fulminant pulmonary edema following IV administration of RCM. The study showed a slight decrease of several C components (C3, C4, and factor B) and a transient consumption coagulopathy. Of particular interest in the latter paper, the authors presented direct hemodynamic and laboratory evidence for pulmonary capillary leakage as the underlying cause of edema, a process known to arise as a consequence of C activation–related sequestration of granulocytes and platelets in the pulmonary microcirculation [28–30]. Another case report postulated that C activation was responsible for the death of a patient undergoing IV pyelography with diatrizoate [10]. Consistent with the key role of anaphylatoxin liberation, the autopsy showed a picture typical of acute respiratory distress syndrome (ARDS), including the presence of granulocytic aggregates in the pulmonary microcirculation [10]. As is known, C activation plays a key role in the development of ARDS [31–36]. Among the more extensive clinical studies examining the role of C activation in RCM reactions, Small et al. [37] analyzed HSRs and C activation in 220 patients undergoing IV pyelography. Nineteen percent of patients displayed HSRs, whereas depressed serum CH50 levels, indicating C activation, occurred in 49%. The RCM-induced decline of CH50/ mL was apparent within 90 seconds after the start of the infusion and returned to normal after approximately 30 minutes. This study highlighted an important fact regarding the rela-

tionship between C activation and HSRs, namely, that more people display signs of C activation reactions than HSRs. Hence, C activation might be present in patients without a clinically manifested reaction, suggesting that anaphylatoxin liberation does not necessarily cause HSRs. C activation might, therefore, be a precondition, or contributing factor to HSRs, but it does not solely explain the phenomenon. Other factors or preconditions might also need to be present in people who develop HSRs. This point was reinforced in the studies by Westaby et al. [38], who demonstrated significant elevation of the anaphylatoxin C3a in the peripheral blood of seven of 11 patients receiving RCM for coronary angiography. In three of seven patients, C3a was increased between four- and 10-fold, but only one of these patients developed symptoms, which were mild. It should be noted that C activation has not been a consistent finding in all clinical studies reporting C measurements in patients given RCM. Kolb et al. [20], for example, found no significant changes in CH50 and hemolytic C3 activity in serum samples obtained from 40 patients before and 30 minutes after undergoing IV pyelography with methylglucamine diatrizoate or iothalamate.

Modulation of Enzymes and Proteolytic Cascades in Plasma One major confounding factor in assessing the role of C activation in RCM reactions is that C is not the only homeostatic plasma enzyme system in blood that is activated during these reactions. It has been described that RCM reactions correlate with diminished concentrations of C1INH in plasma, a protein that inhibits the activation of all four proteolytic cascades in blood—that is, the C, coagulation, fibrinolytic, and kinin-kallikrein systems. This explains the association of HSRs and C activation with accelerated conversion of prekallikrein to kallikrein [10,39,40] as well as the occasional coagulopathy observed in reactor patients. Therefore, one potential cofactor to HSRs might be a proteolytic scission product liberated by the activation of proteolytic chains in blood other than the C system. In fact, some studies claimed that activation of coagulation or the kinin system might play a more important role in the HSRs than C activation, at least in rabbits [41] and guinea pigs [25]. Radiocontrast media–induced, nonspecific proteolysis without participation of the multicomponent C3/C5 convertases might deserve particular attention, because it explains the concurrent activation of coagulation, fibrinolysis, and the kinin-kallikrein systems in humans. This mechanism was proposed by Kolb et al. [20] on the basis that C activation by RCM was not inhibited by ethylenediaminetetraacetic acid (EDTA), which normally suppresses C activation via both the classic and the alternative pathways. The nonspecific lysis resulted in a reduction of CH50 as well as a loss of C4, C2, C3, and C5 hemolytic activities, whereas C6, C8, and C9 were not affected. The parallel

Hypersensitivity Reactions to Radiocontrast Media • Szebeni

production of C3 proteolytic cleavage prouducts and the fact that purified C3 was unaffected by RCM [20] showed that the loss of C3 hemolytic activity was not due to direct alteration of the C3 molecule. Other studies also showed that the C3 and C4 cleavage products that arose as a consequence of RCM-induced, nonspecific proteolysis were different from the C3b and C4b formed during normal C activation, but they formed functional C3 and C5 convertases [10].

29

with universal use of these nonreactogenic contrast media will most likely keep the problem alive for a long time, at least in radiologic centers unable to switch to universal use of these “modern” agents. Future solutions for improved prevention and treatment of RCM reactions include individualized prediction of the reactions—for example, by using in vitro pseudoallergy tests, or application of selective inhibitors of C-mediated HSRs, such as blockers of C activation or action [6••].

Risk Factors, Prevention, and Treatment Radiocontrast media reactions have several risk factors, including tendencies toward allergy (atopic constitution) and asthma, and a history of previous reactions to RCM or other drugs. Severe reactions are largely confined to patients with coronary heart disease [13]. The treatment of acute RCM reactions is largely empiric and supportive. Resuscitation, if necessary, usually involves intubation and infusion of a plasma expander and IV medication with epinephrine, methylprednisolone, diphenhydramine, ranitidine, and atropine sulfate. Vital signs, electrocardiogram, and respiratory functions are continuously monitored until the patient is stabilized. Supportive drugs might be needed for 72 hours [2••,42]. High-risk patients should receive prophylaxis before any diagnostic procedure involving RCM. Treatment of organ dysfunctions induced by RCM reactions might require special attention.

Economic Implications of Radiocontrast Media Reactions Considering that the cost of relatively low-reactogenic, lowosmolarity RCM is substantially higher than that of highosmolarity RCM, some studies addressed the economic sense of switching from universal to selective use of lowosmolarity contrast media. Therefore, Michalson et al. [43] compared the cost of RCM and treatment of adverse reactions with the use of low or high osmolarity RCM. Based on a total of 42,598 radiocontrast studies over 3.5 years, giving an overall reaction rate of 1.2%, the authors pointed out that the incremental cost using low-osmolarity RCM was approximately US$79 per procedure, a figure very close to the estimate of Barrett et al. [13] ($89) a few years before. Because the cost of treating adverse reactions associated with selective use of low-osmolarity RCM was minimal compared with the cost of total conversion to low-osmolarity RCM, the authors concluded that selective use of low-osmolarity RCM meant a savings to their institution of approximately $1 million per year [43].

Conclusions The introduction of low-osmolarity RCM has greatly reduced the threat of HSRs; however, the economic burden associated

References and Recommended Reading Papers of particular interest, published recently, have been highlighted as: • Of importance •• Of major importance 1.• Kumar S, Mahalingam R: Anaphylactoid reactions to radiocontrast agents. Pediatrics 2001, 22:356. This short review touches all major aspects of RCM reactions with updated facts and recommendations. 2.•• Hong SJ, Wong JT, Bloch KJ: Reactions to radiocontrast media. Allergy Asthma Proc 2002, 23:347–351. In the context of a case report, this paper gives a succinct, yet comprehensive, review of RCM reactions. The comparison and classification of available RCMs are particularly useful. 3. Laroche D, Aimone-Gastin I, Dubois F, et al.: Mechanisms of severe, immediate reactions to iodinated contrast material. Radiology 1998, 209:183–190. 4. Courvoisier S, Bircher AJ: Delayed-type hypersensitivity to a nonionic, radiopaque contrast medium. Allergy 1998, 53:1221–1224. 5. Uppal R, Jhaj R, Malhotra S: Adverse drug reactions among inpatients in a north Indian referral hospital. Acad Radiol 1994, 1:59–62. 6.•• Szebeni J: Complement activation-related pseudoallergy caused by liposomes, micellar carriers of intravenous drugs and radiocontrast agents. Crit Rev Ther Drug Carr Syst 2001, 18:567–606. In this review, the mechanism of complement activation by RCM is reviewed in detail. 7.•• Grainger RG: Intravascular radiologic iodinated contrast media. In Diagnostic Radiology: A Textbook of Medical Imaging. Edited by Grainger RG, Allison DJ. Oxford, UK: Churchill Livingstone; 2001:27–41. This textbook provides detailed information on RCMs. 8. Greenberger PA: Contrast media reactions. J Allergy Clin Immunol 1984, 74:600–605. 9. Lieberman P, Siegle RL, Treadwell G: Radiocontrast reactions. Clin Rev Allergy 1986, 4:229–245. 10. Lieberman P: Anaphylactoid reactions to radiocontrast material. Clin Rev Allergy 1991, 9:319–338. 11. Henry DA, Evans DB, Robertson J: The safety and costeffectiveness of low osmolar contrast media. Can economic analysis determine the real worth of a new technology? Drug Saf 1991:28–36. 12. Westhoff-Bleck M, Bleck JS, Jost S: The adverse effects of angiographic radiocontrast media. J Allergy Clin Immunol 1990, 86:684–686. 13. Barrett BJ, Parfrey PS, Vavasour HM, et al.: A comparison of nonionic, low-osmolarity radiocontrast agents with ionic, high-osmolarity agents during cardiac catheterization. Med J Aust 1991, 154:766–772. 14. Katayama H, Spinazzi A, Fouillet X, et al.: Iomeprol: current and future profile of a radiocontrast agent. Invest Radiol 2001, 36:87–96.

30

15.

16.

17.

18.

19.

20.

21.

22.

23.

24. 25. 26.

27.

28.

Anaphylaxis and Drug Allergy

Fattori R, Piva R, Schicchi F, et al.: Iomeprol and iopamidol in cardiac angiography: a randomised, double-blind parallelgroup comparison. N Engl J Med 1992, 326:431–436. Lieberman P: The use of antihistamines in the prevention and treatment of anaphylaxis and anaphylactoid reactions. Singapore Med J 1989, 30:290–293. Younger RE, Herrod HG, Lieberman PL, et al.: Characteristics of diatrizoate-induced basophil histamine release. J Allergy Clin Immunol 1986, 77:94–100. Genovese A, Stellato C, Patella V, et al.: Contrast media are incomplete secretagogues acting on human basophils and mast cells isolated from heart and lung, but not skin tissue. Eur J Radiol 1994, 18(Suppl 1):S61–S66. Napolovlu K, Bolotova EN, Shimanovskii NL: The complement-activating action of modern x-ray contrast agents. J Eksp Klin Farmakol 1998, 61:60–64. Kolb WP, Lang JH, Lasser EC: Nonimmunologic complement activation in normal human serum induced by radiographic contrast media. J Immunol 1978, 121:1232–1238. Vik H, Froysa A, Sonstevold A, et al.: Complement activation and histamine release following administration of roentgen contrast media. Acta Radiol Suppl 1995, 399:83–89. Mikkonen R, Lehto T, Koistinen V, et al.: Suppression of alternative complement pathway activity by radiographic contrast media. Scand J Immunol 1997, 45:371–377. Szebeni J, Baranyi L, Savay S, et al.: The interaction of liposomes with the complement system: in vitro and in vivo assays. Methods Enzymol 2003, 373:136–154. Lasser EC, Sovak M, Lang JH: Development of contrast media idiosyncrasy in the dog. J Radiol 1976, 119:91–95. Lang JH, Lasser EC, Kolb WP: Activation of serum complement by contrast media. Invest Radiol 1976, 11:303–308. Lasser EC, Lang JH, Lyon SG, Hamblin AE: Changes in complement and coagulation factors in a patient suffering a severe anaphylactoid reaction to injected contrast material: some considerations of pathogenesis. Invest Radiol 1980, 15:S6–S12. Vandenplas O, Hantson P, Dive A, Mahieu P: Fulminant pulmonary edema following intravenous administration of radiocontrast media. Acta Clin Belg 1990, 45:334–339. Marceau F, Lundberg C, Hugli TE: Effects of anaphylatoxins on circulation. Immunopharmacol 1987, 14:67–84.

29.

30. 31.

32.

33.

34.

35. 36. 37.

38.

39. 40.

41.

42. 43.

Schmid E, Piccolo MT, Friedl HP, et al.: Requirement for C5a in lung vascular injury following thermal trauma to rat skin. Shock 1997, 8:119–124. Heideman M, Hugli TE: Anaphylatoxin generation in multisystem organ failure. J Trauma 1984, 24:1038–1043. Hammerschmidt DE, Hudson LD, Weaver LJ, et al.: Association of complement activation and elevated plasma-C5a with adult respiratory distress syndrome. Lancet 1980, 8175:947–949. Hohn DC, Meyers AJ, Gherini ST, et al.: Production of acute pulmonary injury by leukocytes and activated complement. Surgery 1980, 88:48–58. Fein A, Wiener-Kronish JP, Niederman M, Matthay MA: Pathophysiology of the adult respiratory distress syndrome. What have we learned from human studies? Crit Care Clin 1986, 2:429–453. Zimmermann T, Laszik Z, Nagy S, et al.: The role of the complement system in the pathogenesis of multiple organ failure in shock. Prog Clin Biol Res 1989, 308:291–297. Bengtsson A: Cascade system activation in shock. Acta Anaesthesiol Scand Suppl 1993, 98:7–10. Kearney ML: Adult respiratory distress syndrome following thoracic trauma. Crit Care Nurs Clin North Am 1993, 5:723–734. Small P, Satin R, Palayew MJ, Hyams B: Prophylactic antihistamines in the management of radiographic contrast reactions. Clin Allergy 1982, 12:289–294. Westaby S, Dawson P, Turner MW, Pridie RB: Angiography and complement activation: evidence for generation of C3a anaphylatoxin by intravascular contrast agents. Cardiovasc Res 1985, 19:85–88. Lasser EC: Adverse reactions to intravascular administration of contrast media. J Allergy 1981, 36:369–373. Hoffmeister HM, Heller W: Radiographic contrast media and the coagulation and complement systems. Invest Radiol 1996, 31:591–595. Lasser EC, Lang JH, Lyon SG, Hamblin AE: Complement and contrast material reactors. J Allergy Clin Immunol 1979, 64:105–112. Oh VM: Treatment of allergic adverse drug reactions. J Allergy Clin Immunol 1986, 77:94–100. Michalson A, Franken EAJ, Smith W: Cost-effectiveness and safety of selective use of low-osmolarity contrast media. Acta Radiol 1994, 1:59–62.